Development of a Dielectrophoresis-Based Cancer-Cell Analysis Tool
Douglas, Temple Anne
MetadataShow full item record
One significant obstacle in cancer treatment is tumor heterogeneity. Different subpopulations within a tumor can respond differently to chemotherapy, resulting in resistance and recurrence. Addressing these differences while choosing a treatment modality could significantly improve chemotherapy outcomes. This work focuses on the development of a new modular device that leverages the unique advantages of a contactless dielectrophoresis, a method that uses applied electric fields in a microfluidic device to separate cells by biophysical phenotype. By optimizing force balancing between the dielectrophoretic force and the drag force on cells in the device, and by using cell-size pillars to maximize electric field gradients per volt applied while reducing cell-cell interactions,we demonstrate that it is possible to separate mouse ovarian surface epithelial (MOSE) cells at different stages while maintaining high viability. We also show other cell types to be separable with this device and develop an algorithm to rapidly analyze cell response to a variety of frequency/voltage/flow rate combinations. We also propose a microfluidic device downstream of the DEP chip that can be used to provide an integrated system for studying the subpopulations separated using dielectrophoresis by moving them into a culture chamber with hydrogel where they can be grown in 3D and characterized for a variety of parameters such as biophysical structure, metastatic capacity, and chemotherapy resistance.
General Audience Abstract
Dielectrophoresis is a method by which cells are polarized in response to an electric field gradient. This work optimizes this technique so that it can be used to separate highly similar subpopulations of cancer cells in a microfluidic device. Computer code is also developed to automate data processing. A technique for analyzing these cell subpopulations is also proposed and some feasibility testing performed.
- Doctoral Dissertations